Action unit classification using active appearance models and conditional random fields.
Laurens van der MaatenEmile A. HendriksPublished in: Cogn. Process. (2012)
Keyphrases
- conditional random fields
- information extraction
- hidden markov models
- structured learning
- classification accuracy
- random fields
- probabilistic model
- structured output
- text classification
- graphical models
- markov random field
- generative model
- pattern classification
- supervised learning
- maximum entropy
- machine learning
- named entity recognition
- information retrieval
- pairwise
- decision trees
- feature selection
- higher order
- sequence labeling
- semi markov
- support vector machine
- web page prediction
- approximate inference
- parameter learning
- structured prediction
- protein fold recognition
- error correcting output coding
- cost sensitive
- class labels
- machine learning algorithms
- training samples
- active learning
- feature vectors
- training set